At Kairos, one of our main goals is to better understand how developers use our products. Knowing your challenges and successes helps us learn from each other but also sharing those challenges and successes helps us all connect in our community.
Each month we'll be featuring an interview with our customers allowing them to share their journey, knowledge, and advice with our API and SDKs.
This is our second post in our “Developer Discussions” interview series and we are proud to share this story. Meet Emma Yang, a developer who has created Timeless, an app designed to help Alzheimer’s patients recognize their family and friends through facial recognition. One of the most amazing facts about Miss Yang is that she is just 13 years old. Yes, you read that correctly - she is 13.
Meet the Developer
Going Forward: EY will stand for Emma Yang
Tell us a little bit more about yourself. Example: Who are you, what do you do, why are you into programming, and what did you do with our technology etc;?
EY: I’m a 13 year old Middle School student in New York City. I am passionate about biotech, computer science, machine learning, and technology innovation. I am also especially interested in the intersection between health care and computer science, and to apply my interest to create solutions that can solve real world problem.
Out of all the cutting-edge technologies, Machine Learning and Artificial Intelligence fascinates me the most. Hence when I was contemplating creating an app, I started to think about how I can leverage ML and AI technology to make a wide impact for the community.
I am also a believer of encouraging girls to learn STEM and Computer Science. I founded my school’s first Coding and Innovation Club for middle school and I’m currently running a coding workshop to teach middle school girls how to code [through] programs like Technovation. (I am a Regional Alumni Ambassador of Technovation. I went to Technovation in 2015 and very fortunately won first place in the US and 2nd place globally).
I came across Kairos when I was looking for a facial recognition technology for my app Timeless. Kairos was (and is) the perfect solution because it was easy to integrate and it has the ability to recognize a face within a photo with highest accuracy. The technology was used to help Alzheimer’s patients to identity and recognize their loved ones.
Can you tell us a little bit about Timeless and how you came up with this idea?
EY:Timeless is a simple and easy to use mobile app for Alzheimer’s patients to remember events, stay connected and engaged with friends and family, and to recognize people through automatic facial recognition technology.
My Grandma has Alzheimer’s Disease [and] the illness made her forget many things that were very important to her. We stay connected by sending her photos but she doesn’t always remember who the people in the pictures are because recognizing people and remembering who they are is becoming increasingly difficult. She needs her caregiver to help her remember to do everyday tasks such as taking medication. Her illness is preventing us from staying connected with each other.
After doing extensive market research, I found out that none of the apps on the market related to Alzheimer’s disease are targeted for the patients themselves. The available apps only targeted caregivers seeking help or people who want to learn more about the disease. I wanted to leverage the power of artificial intelligence and mobile technology to help Alzheimer’s patients like my Grandma stay connected with friends and family.
How did you find out about Kairos and our technology?
EY: Early last year, when I started to develop the prototype of Timeless, I was looking for platforms for facial recognition. For Timeless, I needed a facial recognition service that could recognize who a person is in a photo rather than recognizing whether there is/are person(s) in it.
From my research, I came across services such as Google Cloud Vision API, Microsoft Cognitive Services, IBM Watson Visual Recognition, Lambda Labs, and Kairos. There are also lower level platforms such as OpenCV, but I would need to do a lot of coding, neural network training, and configuration myself.
Kairos is the only platform provider which provides an easy to use API for what I need. Other services which are capable of recognizing identities of persons from Apple, Google, and Facebook do not provide an API for developers to tap into.
Which tool of ours did you use and why?
EY: We wanted to enjoy this hackathon and do something fun and different. Jibo seemed like fun and new. And also the idea to play a theme song when you come to office immediately clicked with me. (Seth came up with the idea)
I am using the Facial Recognition service provided by Kairos [as] this service provides a RESTful API which can be easily integrated into a mobile app. I tried out the API in the test area on the Kairos website. Not only is it easy to configure and use, after enrolling a single photo of a person, it can very efficiently recognize and identify the person from subsequent photos and with very high accuracy. This efficiency and speed is critical for providing a meaningful and useable functionality for applications like Timeless.
Can you explain to us the process in which you created your application and how our technology fit into it?
EY: After having the idea of developing an app for the Alzheimer's patients, I conducted research to understand more about the severity of the illness; how many people are impacted, what are the symptoms, and how it affect the patient’s everyday life? I then researched what tools or applications are out there which are directed to patients.
For the functions and features of the app I worked with Dr. Melissa Kramps, a specialist in the Alzheimer's Disease & Memory Disorders Program at the New York-Presbyterian Weill Cornell Medical Center, to validate and seek suggestion on areas the app should focus on.
On the technical side, I reached out to Cole, [CTO of Kairos], for sponsorship and technical guidance on how to best leverage and integrate Kairos [facial recognition].
For the development phase, I created flowcharts to work out the workflow and storyboards for initial UI design. I then used UI design tools, [like] Sketch, to detail the UI design [and] prototyping tools, [like] Flinto, for the workflow. After getting these aspects right, I coded the working prototype in Xcode using Swift.
What challenges did you face while building the application? Any success (examples may be: scholarships, investors, or even the ease of the integration of the technology etc)?
EY: The biggest challenge was the display of the result of the facial recognition on the app was very slow in the beginning, even though the Kairos API was responding within a few seconds. This would be a [problem] for the application when patients are using it. After a lot of debugging, I found out that the problem was with the third party library (SwiftyJSON/Cocoapods) that I was using to process the data returned by Kairos. This problem was solved after I switched to the native library provided by Apple. There was also an issue with the performance with screen transition, which I optimized. After I fixed these two problems the turnaround time went from over 60 to around 5 seconds, a significant improvement. The performance further improved to about 3 seconds after the upgraded Kairos service was used.
I’m very fortunate that the idea and potential impact of the app has been recognized by a number of media outlets:
- BBC World Service
- Alzheimer’s Disease Research
- Senior Planet
- Next Avenue
- Star 2 (Malaysia)
- El Mundo (Spain)
What’s next for Timeless? Do you plan on further refining it?
EY: I am planning to take Timeless from the prototyping phase into a real product, so that patients with Alzheimer's can use the app to connect to their friends and family. At this stage, I’m focusing on developing the iOS version, later I would like to expand [to other devices and platforms] tablets, and also add multilingual support to increase its reach.
I am also interested in adding voice interfaces such as Amazon Alexa or Apple’s Siri to further expand the use case and enhance the ease of use of the product. The capability of Kairo’s facial recognition in videos is also something that I would like to leverage to further build out the functionality of Timeless.
What are you working on now?
EY: I just started to work with a UI/UX designer to revamp the UI and workflow and make the app look more professional and more suitable for elderly users. I will also soon start to develop the back end of the app, likely using Google Firebase. I am also planning to crowdfund for the development, hosting and marketing costs.
I’m also a team member of the robotics team in my school and we are working on constructing and programming a robot for the 2017 FIRST Robotics Competition.
Separately, I’m also working on project for identifying cancerous lung CT Scan images using AI and Machine Learning using Wolfram Mathematica.
How do you feel/what do you think about human analytics (facial recognition and emotion analysis specifically) integration with technology?
EY: I think human analytics (facial recognition and emotion analysis) are fascinating technologies. The potential benefit when integrated into devices and environments is huge. Imagine we can have smart home, or even smartphones that can recognize who you are and what your mood is and act or recommend actions accordingly, or facilities or even vehicles where occupants can be monitored to see if they are not feeling well. Such technologies can add a lot of “depth” to the smart devices or environments by understanding their user in a much deeper level, hence expanding their usefulness by orders of magnitudes.
It is impressive how young and talented you are. How did you get into coding and what resources helped you learn more about the field?
EY: My Dad is a software engineer, he introduced me to coding when I was 8 years old. I started with Scratch, a block-based programming language for kids, and I found it fascinating. Afterwards I learned HTML/CSS web development, Java, Object Oriented Design, MIT App Inventor, LEGO Mindstorm Robots programing, and iOS mobile development in Swift. Timeless is the first mobile app I built from the ground up.
To learn about coding, I mostly use online resources and courses (such as the CTY program developed by Johns Hopkins), I also attended workshops and Summer camps for robotics and iOS. In the Wolfram Mentorships Program I learned the Wolfram Language and did a few projects such as NYC Traffic Collision Analysis, Finding Summer Meal Sites using Open Data and took part in the Whitehouse’s Opportunity Project last year.
What do you think you will be doing 10 years from now?
EY: I would like to study computer science, artificial intelligence, machine learning, and biotechnology by the time I go to college in a few years. I hope that I would be able to use AI and Machine Learning to create things that make a difference in people’s lives.
I hope by then Timeless has already reached and helped millions of Alzheimer’s patients and caregivers worldwide. I hope by then there is already a cure for the disease.
I also hope I will have found a company that creates really cool and amazing products that people love. And that equality would have been reached for girls in STEM.
The Future is Bright
It’s no doubt that Miss Yang is one of the many bright minds of tomorrow as her list of accomplishments is long and growing. She, and other young coders, will define our future tomorrow so keeping an eye on her, and others, will give you a view into what technological advancements the younger generations will bring to the table. When asked for advice for future coders, Yang ment business “Never let people tell you it is not possible just because you are young or you are a girl”. To learn more about Timeless and Miss Yang you can visit these sites: